Summary The PA-31-310, registrationC-FARL, serial number31306, operated by Les Ailes de Gasp inc., with one pilot and two passengers on board, was on a visual flight rules flight from les-de-la-Madeleine, Quebec to Gasp, Quebec. While en route to Gasp, the pilot was informed about weather conditions at his destination, which were a ceiling at 500feet and visibility of mile in fog. The pilot requested clearance for an instrument approach, which he received at approximately 1857 eastern daylight time. A few seconds later the pilot switched on the aerodrome lights with his microphone button. That was the last radio transmission received from the aircraft. When the aircraft did not arrive at its destination, emergency procedures were initiated to find it. The wreckage was found the next day at 1028 eastern daylight time on a hilltop 1.2nautical miles (nm) north-east of the airport. The aircraft was destroyed, but did not catch fire. The three occupants were fatally injured. Ce rapport est galement disponible en franais. Other Factual Information The PA-31 had been chartered to transport one passenger from Gasp to les-de-la-Madeleine, then return to Gasp with two passengers. Before departing Gasp, at approximately 1612 eastern daylight time,1 the pilot visited the NAV CANADA Web site for a weather report. The terminal aerodrome forecast (TAF) for Gasp issued at 1930 coordinated universal time (UTC) was as follows: between 2000and 0800UTC, scattered cloud at 800feet above ground level (agl), ceiling at 3000feetagl, visibility over 6miles; and temporarily between 0000and 0800UTC, ceiling 800feetagl. The cloud and weather chart for the graphic area forecast (GFA), valid from 1800UTC, indicated the possibility of a ceiling at 200feetagl and fog patches, reducing visibility to mile along the shores of the Gulf of St.Lawrence. The pilot arrived at Gasp airport around 1645. He seemed to be in good condition and ready for the flight. He filed a visual flight rules (VFR) flight plan for the return trip. The aircraft's four fuel tanks were filled, allowing him to complete the flight as planned in accordance with existing regulations. The aircraft took off at approximately 1705 with an anticipated return time of approximately 1845. The flight to les-de-la-Madeleine was without incident, and the aircraft landed there at approximately 1800. Twelve minutes later, with two passengers on board, the aircraft took off for Gasp. While the aircraft was en route, the TAF for Gasp was revised twice, at 1839 and again at 1849. These two revisions indicated deteriorating weather conditions compared to the TAF received prior to departure; the initial ceiling forecast of 800feetagl dropped to 300feetagl, and the forecast for visibility was mile in fog. There is no evidence that the pilot either requested or was advised of these revisions. At 1849:58, the pilot tried to make initial radio contact with the Flight Service Station (FSS) at Qubec. He transmitted twice on 122.3megahertz (MHz), but received no response. The pilot then tried to make contact on 126.7MHz. The FSS specialist responded and asked the pilot to switch back to 122.3MHz. When two-way communication was established on 122.3MHz at 1853:32, the pilot was advised that the surface winds were favourable for runway11, and the altimeter setting was 30.25inches of mercury (in.Hg). The pilot read back the altimeter setting correctly and advised that he would proceed for runway11. At 1855:13, the FSS specialist gave the pilot the latest weather observation from Gasp, which was a special bulletin issued at 2241 UTC. It indicated a ceiling at 500 feet agl and visibility of mile in fog. Based on this information, the pilot advised that he would proceed for runway 29, but did not specify the type of approach. At approximately 1856:07, when he was about 7nm south-east of Gasp, the pilot requested clearance for an instrument approach, which he received less than one minute later. At 1857:20, the pilot pressed his microphone button seven times to switch the aerodrome lights on high. That was the last radio transmission received from the aircraft. According to the information received, all lights were working normally at the time of the occurrence. Except for the call made when 7nm south-east, the pilot made no reports during the approach. When the aircraft did not arrive at its destination, emergency procedures were initiated. The emergency locator transmitter (ELT), modelTEL82, serial number10246, was installed and maintained in accordance with the regulations. It was found with its selector switch in the automatic position, but it did not activate because the battery disconnected on impact. This delayed locating the aircraft until 1028 the next day. Tests in the Transportation Safety Board (TSB) Engineering Laboratory showed that the ELT operated normally when the battery was connected. Although they were wearing their seat belts at the time of the accident, all three occupants were thrown from the aircraft with their seats due to the force of the impact. There was no chance of survival because of the force of the impact. The aircraft crashed on the summit of a hill with an elevation of about 300feet above sea level (asl), 1.2nm north-east of the threshold of runway29, and 0.8nm north of the approach track. The swath cut through the trees by the aircraft extended over a distance of about 100metres and followed a trajectory leading directly towards the runway. The debris pattern at the crash site indicated a high-speed, low-angle impact. Marks left on one of the speed indicators indicated a speed of 185miles per hour (mph) on impact, which is far greater than the normal approach speed of 110mph. All flight control surfaces were retrieved at the site, and all damage to the aircraft was attributed to impact forces. The flaps were retracted, and the landing gear was not in the down and locked position. Several aircraft instruments, including the altimeters, were destroyed on impact and could not be analysed. The aircraft records indicate that both altimeters were calibrated in June2003, and no deficiencies had been reported since then. Examination of the propeller hubs revealed that the engines were producing power at the time of impact. Engine rpm could not be determined. Examination of the airworthiness directives, service bulletins, and technical log books of the aircraft indicate that it was certified, equipped, and maintained in accordance with existing regulations and approved procedures. There was no evidence found of any airframe failure or system malfunction prior to or during the flight. The weight and centre of gravity of the air plane were within the limits prescribed by the manufacturer. The pilot held a valid airline transport pilot licence with an instrument flight (IFR) endorsement. The last entry in his flight log book indicates that as of 16December2002 he had 5262flying hours, of which 3500hours were instrument flight. He was qualified on the PA-31 and had over 3000hours on this type of aircraft. The pilot was from the Gasp region, where he had gained most of his flying experience. He was president of the company, as well as chief pilot and director of operations. According to the autopsy and toxicology testing, there was no indication that physiological factors affected the pilot's performance. Staffing at the Qubec FSS was adequate on the evening of the accident. The FSS specialist in charge of the Gasp sector had been on duty since 1115, and his workload was considered light. All necessary equipment was in good condition and was being used. To provide a common channel for air-ground communications with aircraft operating in remote areas of Canada, FSS specialists monitor frequency 5680kHz on the high frequency (HF) band. On the evening of the accident, it was reported that the HFband was producing radio interference. It is recognized that HFbands produce this type of interference. It is possible that this interference prevented the FSS specialist from hearing the first messages transmitted by the pilot on 122.3MHz. Since the Gasp airport is outside controlled airspace, NAV CANADA does not provide radar coverage for aircraft flying at low altitude in this area. The only data obtained came from military radar and covered only a short segment of the flight when the aircraft was in cruise flight about 30miles south-east of Gasp. Those data indicate that the aircraft was heading directly for the airport at normal cruise speed and suggest that the pilot was using the global positioning system (GPS) to navigate and report his distance from the airport. The GPS installed on the aircraft, a Garmin GPSMAP295, was not certified for instrument navigation, but could be used to facilitate visual navigation. The condition of the GPS after the accident did not allow investigators to obtain any information useful for the investigation. Regulations permit the aircraft to conduct instrument flights with passengers on board without a co-pilot provided that it is equipped with an autopilot. Although using an autopilot reduces the pilot's workload, the presence of a co-pilot allows tasks to be shared and offers a better opportunity to detect deviations from the desired flight profile. Examination of the autopilot control console did not reveal whether or not it was in operation prior to or at the time of impact. It was not required to be in operation. The aircraft was not equipped with a ground proximity warning system (GPWS) or a radio altimeter, nor were they required by regulation. The published minimum descent altitude (MDA) for the runway29 back course is established at 440feet asl and a visibility of one mile. The elevation of the aerodrome is 108feet asl. Even if the reported visibility was less than the minimum published for an instrument approach, the pilot was not prohibited by regulation from conducting the approach. With regard to the landing, the existing regulations prohibit the pilot of an aircraft on an instrument approach from continuing the descent below the MDA if he does not establish and maintain the visual reference required to land safely. If the pilot loses the required visual references, he must execute a go-around. The aircraft was found over 25degrees to the right of the localizer track. The design of the approach system at Gasp, which dates from the1970s, and the surrounding hilly terrain contribute to signal interference in the area between 25and 35degrees from the localizer. This interference causes the course deviation indicator (CDI) to oscillate when an aircraft is in this area, contrary to the standard established by the International Civil Aviation Organization, which states that the signal should allow the CDI to remain stable on the appropriate side. To inform pilots of this signal interference, a note on the approach plate indicates that the Gasp localizer transmitter is not reliable beyond 25degrees on either side of the localizer. In-flight tests have shown that despite the oscillation of the CDI inside this area, the indication remains appropriate in relation to the localizer track. After the accident, an in-flight test of the navigation aids at Gasp was conducted, and it showed that they met operational requirements and that broadcast parameters were within technical tolerances. No navigation aid malfunctions were reported on the day of the accident. On 16 December 1997, a CL-600-2B19 crashed at the Fredericton (NewBrunswick) airport while executing a go-around in reduced visibility and low ceiling conditions. The TSB investigation of this accident (reportA97H0011) identified 28other accidents in Canada between 01January1984 and 30June1998 involving heavy aircraft landing in reduced visibility conditions where these conditions contributed to the accident. This investigation also identified a safety deficiency due to the fact that the existing regulations do not provide sufficient protection against the risk of collision with the terrain when instrument approaches are conducted in reduced visibility conditions. In its report, published on 20May1999, the TSB recommended that: The Department of Transport reassess CategoryI approach and landing criteria (re-aligning weather minima with operating requirements) to ensure a level of safety consistent with CategoryII criteria. Transport Canada responded to the recommendation on 06 August 1999 indicating that a draft regulation amendment to strengthen the standards applicable to instrument approaches in minimal weather conditions would be submitted without delay to the Canadian Aviation Regulation Advisory Council (CARAC) for comment with the objective of applying the changes as soon as possible. On 12 August 1999, a Raytheon Beech1900D crashed on approach to the Sept-les, Quebec airport, when the reported weather conditions indicated a ceiling of 200feet and a visibility of statute mile. The TSB investigation into this accident (reportA99Q0151) identified four other accidents that had occurred with reduced visibility as an underlying factor since recommendationA99-05 had been issued. The TSB report on this accident, published 14March2002, included a Board recommendation that: The Department of Transport expedite the approach ban regulations prohibiting pilots from conducting approaches in visibility conditions that are not adequate for the approach to be conducted safely. Transport Canada responded to the recommendation on 26 May 2002 indicating that they had prepared 16notices of proposed amendment (NPA2000-001, 002, 006, 007, 008, 009, 010, 011, 012, 106, 107, 108, 116, 117, 194, and195) to address the issue of a regulatory approach ban related to visibility. The response stated that the NPAs were, at the time, under review by the Department of Justice and that the final version should be published in the Canada Gazette in June2002. Information received recently from Transport Canada revealed, after some delay, that the NPAs are now in the hands of the Minister of Transport for approval. Previously, from 2000to July2004, the NPAs were at the Department of Justice (filenumber10000-386) for revision of the wording and review of legality. This prolonged delay was apparently due, in part, to the high priority that the CARAC (composed of senior safety and security managers) accorded to the treatment of the draft security regulations following the events of 11September2001. Increased demand for the services of the Department of Justice at the time apparently caused additional delays. Since recommendationA02-01 was issued, TSB has identified another accident, in addition to the one that is the subject of this report (reportA03Q0151), where visibility was an underlying factor. On 25February2004, at Edmonton International Airport, a B737-200 touched down beside the runway during a landing in which weather conditions were a ceiling at 300feet and visibility of 1/8statute mile in fog, and the runway visual range (RVR) was 1200feet. The aircraft sustained substantial damage. Fortunately, none of the 36occupants were injured (TSB occurrenceA04W0032; investigation in progress).